Colorectal cancer (CRC) kills 50,000 Americans every year. Half have hepatic metastases at diagnosis;only a small percentage can be cured. Our overriding goal is to develop new tools to deliver therapeutic genes to hepatic CRC metastases and new tools to monitor non-invasively therapeutic gene delivery and efficacy. CRC tumors often overexpress both carcinoembryonic antigen (CEA) and cyclooxygenase 2 (COX-2). Hepatic metastases generally express even greater levels of CEA and COX-2 than their primary tumors. The aims of our currently funded Research Component 4 project are to: (1) Establish 00X2"" CEA* CRC hepatic xenografts that can be monitored by non-invasive imaging, as models of hepatic CRC metastasis. (2) Create AdCox2fLucTK, an adenovirus from which a firefly luciferase/HSVI-TK fusion reporter-therapeutic protein is regulated by the Cox2 promoter. (3) Characterize sCAR-aCEAs, adenovirus liver-untargeting/CEA* tumor-retargeting agents that fuse the Coxackie and Adenovirus Receptor (CAR) extracellular domain to a single-chain anti-CEA antibody. (4) Examine the ability of (i) sCAR-aCEA reagents to retarget Ad gene delivery vectors to CEA* CRC hepatic xenografts and the ability of (ii) Cox2 regulatory elements in Ad vectors to restrict transgene expression to COX-2* CRC xenografts, and (5) Optimize ganciclovir-dependent killing of hepatic C0X2* CEA* CRC metastases by intravenously administered, Cox2 transcriptionally restricted, transductionaily re-targeted [Ad.Cox2fLucTK][sCAR-aCEA]. We accomplished all these goals. This project received the lowest (best) priority of the current ICMIC Research Component projects. However, several criticisms from the Summary Statement form the basis of this renewal application. What are: (1) The effects of innate immunity on transductional retargeting and transcriptional restriction? (2) The effects of preexisting immunity on Ad transductional retargeting and transcriptional restriction? (3) The effects of shed, circulating soluble CEA on Ad retargeting? (4) The final criticism was a lack of a plan for clinical translation. Our Specific Aims are: (1) To establish C0X2* CEA* hepatic CRC models in immunocompetent B57BI6 and CEA-transgenic mice, and to use this model to (2) minimize the innate immune response to transcriptionally restricted/transductionally retargeted Ad imaging and therapeutic vectors, (3) optimize Ad vector evasion of neutralizing anti-ad antibody, and (4) optimize inhibition of sCAR-aCEA retargeting of Ad vectors to CEA* hepatic CRC metastases by shed CEA. (5) Our final aim is to design, develop reagents for, obtain approval for and initiate a clinical trial of Ad.Cox2HSV1sr39tk/FHBG PET imaging of CRC metastases in patients.